Icarus 191 (2007) 486–496
www.elsevier.com/locate/icarus
The lunar exosphere: The sputtering contribution
P. Wurz a,∗ , U. Rohner a,1 , J.A. Whitby a , C. Kolb b , H. Lammer b ,
P. Dobnikar c , J.A. Martín-Fernández d
a Physikalisches Institut, Universität Bern, Sidlerstrasse 5, CH-3012 Bern, Switzerland
b Space Research Institute, Austrian Academy of Sciences, Schmiedlstrasse 6, A-8042 Graz, Austria
c Institute for Physics, University of Graz, Universitätsplatz 5/II, A-8010 Graz, Austria
d Department for Computer Science and Applied Mathematics, University of Girona, Edifici, P-IV, Campus Montilivi, E-17071 Girona, Spain
Received 12 February 2007; revised 26 April 2007
Available online 24 May 2007
Abstract
We have extended our Monte Carlo model of exospheres [Wurz, P., Lammer, H., 2003. Icarus 164 (1), 1–13] by treating the ion-induced
sputtering process from a known surface in a self-consistent way. The comparison of the calculated exospheric densities with experimental data,
which are mostly upper limits, shows that all of our calculated densities are within the measurement limits. The total calculated exospheric density
at the lunar surface of about 1 × 107 m−3 as result of solar wind sputtering we find is much less than the experimental total exospheric density of
about 1012 m−3 . We conclude that sputtering contributes only a small fraction of the total exosphere, at least close to the surface. Because of the
considerably larger scale height of atoms released via sputtering into the exosphere, sputtered atoms start to dominate the exosphere at altitudes
exceeding a few 1000 km, with the exception of some light and abundant species released thermally, e.g. H2 , He, CH4 , and OH. Furthermore,
for more refractory species such as calcium, our model indicates that sputtering may well be the dominant mechanism responsible for the lunar
atmospheric inventory, but observational data does not yet allow firm conclusions to be drawn.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Moon; Moon, surface; Atmospheres, composition
1. Introduction
Rather surprisingly, the composition and structure of the tenuous lunar atmosphere, actually an exosphere, remain poorly
understood almost forty years after the first Apollo landings,
and as recently as 1997 it has been suggested that we cannot account for 90% of the night-time exosphere (Stern et al.,
1997, and references therein). The reasons for this are the difficulty of observations due to the very low number densities,
and the complexity of models due to the multiplicity of mechanisms responsible for the input and loss of atomic species to
and from the exosphere. These mechanisms include ion sputtering, photon stimulated desorption (PSD) and micro-meteoroid
impact vaporization resulting in inputs to the atmosphere, and
* Corresponding author. Fax: +41 31 631 44 05.
E-mail address: [email protected] (P. Wurz).
1 Present address: Lawrence Livermore National Laboratory, Physics and Ad-
vanced Technologies Directorate, Livermore, CA 94551, USA.
0019-1035/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.icarus.2007.04.034
photoionization, surface adsorption and thermal escape resulting in losses from the lunar gravity field (e.g., Stern, 1999;
Killen and Ip, 1999; Hunten and Sprague, 1997; Mendillo et
al., 1999).
Within the last twenty years, much work has focused on the
neutral sodium and potassium components of the exosphere
as these can now be studied from the Earth (Potter and Morgan, 1988, 1998; Flynn and Mendillo, 1993), but the behavior
of these elements may not be representative of other species
as these are the elements probably most influenced by meteoritic influx and photon-induced desorption (Sprague et al.,
1992). The composition of noble gases in the lunar atmosphere
(largely inferred from studies of gas trapped in lunar regolith
samples) indicates that these may be dominated by a solar wind
source, but with additional contributions probably from the interior of the Moon (e.g., Hodges and Hoffman, 1975; Wieler
et al., 1996). The latter idea is supported by observations of
episodic outgassing of radon (Gorenstein et al., 1974a, 1974b;
Hodges, 1975; see Lawson et al., 2005, for a review of the liter-
The lunar exosphere: The sputtering contribution
487
Table 1
Observational upper limits for lunar neutral elemental exosphere dayside number densities (ignoring hydrogen and noble gases) after Stern (1999)
Species
Detection method
Number density (cm−3 )
Reference
O
N
C
S
Na
Si
Al
Ca
K
Fe
Ti
Ba
Al
Mg
Apollo 17 UV spectroscopy (1304 Å)
Apollo 17 UV spectroscopy (1200 Å)
Apollo 17 UV spectroscopy (1657 Å)
Apollo 17 UV spectroscopy (1474 Å)
Ground-based spectroscopy (5889 Å)
Ground-based spectroscopy (3906 Å)
Ground-based spectroscopy (3962 Å)
Ground-based spectroscopy (4227 Å)
Ground-based spectroscopy (7699 Å)
Ground-based spectroscopy (3859 Å)
Ground-based spectroscopy (5036 Å)
Ground-based spectroscopy (5536 Å)
HST UV spectroscopy
HST UV spectroscopy
<500 (3σ )
<600 (3σ )
<200 (3σ )
<150 (3σ )
35,70
<48 (5σ )
<55 (5σ )
<1 (5σ )
17
<380 (5σ )
<1 (5σ )
<0.2 (5σ )
<11,000 (5σ )
<6000 (5σ )
Feldman and Morrison (1991)
Feastie et al. (1973)
Feldman and Morrison (1991)
Feldman and Morrison (1991)
Potter and Morgan (1988, 1998)
Flynn and Stern (1996)
Flynn and Stern (1996)
Flynn and Stern (1996)
Potter and Morgan (1988)
Flynn and Stern (1996)
Flynn and Stern (1996)
Flynn and Stern (1996)
Stern et al. (1997)
Stern et al. (1997)
Note. Data from Apollo 17 are derived from averages of different observation positions.
ature). Because the solar wind impinges on the lunar surface H
and He atoms are neutralized and are subsequently released to
become part of the exosphere (e.g., Hinton and Taeusch, 1964;
Johnson, 1971; Hodges, 1973, 1980). In contrast, the flux of
heavier more refractory elements to the lunar atmosphere may
be dominated by ion sputtering. In this case, the lunar surface
material will be the primary source reservoir for elements in the
lunar exosphere such as Si, Ti, Al, Fe, Mg, Ca, and O, which
have not yet been observed directly. Table 1 shows upper limits
for the abundances of these and some other elements. The exosphere densities at the surface of Na and K given in Table 1 are
not only the result of sputtering but also to a larger extent most
likely the result of photon stimulated desorption (PSD). In addition it has been estimated that up to 15% of the observed Na in
the exosphere can be attributed to the impact of micro-meteors
(Flynn and Mendillo, 1993).
Hodges (1973, 1980) modeled global distributions of H and
He concentrations in the lunar exosphere and got results for He
that are in good agreement with the Apollo 17 lunar surface
mass spectrometer measurements. While in view of the absence
of hydrogen the author proposed a model where the amounts are
assumed to be below the measurement threshold and the majority of the hydrogen, which is of solar origin, is efficiently lost
from the Moon by thermal and nonthermal escape processes.
In the present work we investigate the contribution to the lunar
neutral exosphere of heavy refractory elements released by ion
sputtering.
A quantitative understanding of the processes that give rise
to the lunar exosphere in a self consistent way will improve
the general understanding of the exospheres of atmosphere-less
bodies. For airless bodies the exosphere is directly connected
to the planetary surface. Lunar exosphere measurements with
the SARA instrument on the Indian Chandrayaan-1 mission,
scheduled for launch in early 2008, will allow the testing of our
model calculations (Bhardwaj et al., 2005). An important future application of models of exospheres of airless bodies will
be the Hermean exosphere where measurements of its composition and structure will be performed by several instruments,
to gain information not only about the exosphere itself but also
about the surface composition of Mercury (Milillo et al., 2005).
Our approach is to test our models against constraints obtained
by current and future measurements of the lunar exosphere,
and then to extend the models to the slightly different conditions relevant on Mercury. Although our modeling approach is
quite general, we will concentrate on the predictions made for
species where sputtering is expected to dominate the production
rate. We do this first because it will allow a better check of the
model (when good enough observational constraints are available), and secondly because the abundance of several of these
species (e.g., Ca, Fe) is critical to constraining models for the
formation and subsequent evolution of the Moon and especially
Mercury.
2. Lunar surface composition
The topmost layer of the lunar surface consists largely of
regolith, an unconsolidated “soil” layer of several meters thickness, made of lithic fragments, agglutinate grains and dust, continuously reworked and turned over by meteorite bombardment
over geologic timescales (e.g., Chapter 7 in Heiken et al., 1991).
This bombardment results in the introduction of some meteoritic material of less than 2 wt% in soils (Papike et al., 1998)
and some local mixing of the surface. Laul and Papike (1980)
were able to explain the composition of soil samples using between three and six end-members for each Apollo site, although
the choice of end-members was adjusted for each site. In this
work we have used measured regolith compositions reported
by Papike et al. (1982) for the Apollo and Luna soil samples.
These can be divided into the following categories: Highland,
Ti-rich Mare, Ti-poor Mare and Kreep. These names reflect the
dominant petrological component in these soils deduced using
the same component analysis as Laul et al. (1982) and are between them believed to be broadly representative of the lunar
surface. To simplify model calculations we simply averaged the
values in each category (see Table 2), effectively giving a fourcomponent model for the composition of lunar soils. Although
the composition of the finest soil fractions differs from the bulk,
we have assumed that the average bulk composition is repre-
488
P. Wurz et al. / Icarus 191 (2007) 486–496
Table 2
Lunar reference suite soils (including Apollo 15 sample 15601) for Highland, Kreep, low-Ti and high-Ti Mare regions
Reference suite
Size fraction
Si
Ti
Al
Fe
Mg
Ca
Na
K
Mn
Cr
O
Total
Highland soils
64501,122
67461,74
22001,35
72501,15
Bulk
Bulk
>125 µm
Bulk
16.26
16.09
∗16.18
16.57
0.10
0.08
0.13
0.39
11.72
12.31
9.95
8.69
1.26
1.26
2.18
2.91
2.62
2.08
5.19
5.47
6.61
6.74
5.43
4.91
0.31
0.30
0.24
0.31
0.05
0.03
0.03
0.08
0.02
0.02
0.03
0.04
0.03
0.02
0.05
0.07
61.03
61.09
60.59
60.57
100
100
100
100
16.31
0.17
10.66
1.90
3.84
5.92
0.29
0.05
0.03
0.04
60.82
100
17.21
17.47
17.37
0.79
0.64
0.44
5.51
6.23
7.70
5.38
4.80
3.22
5.80
5.11
5.26
4.37
4.43
4.49
0.35
0.48
0.50
0.12
0.19
0.26
0.07
0.06
0.04
0.12
0.11
0.06
60.29
60.47
60.66
100
100
100
17.35
0.62
6.48
4.47
5.39
4.43
0.44
0.19
0.06
0.10
60.47
100
17.21
17.39
∗16.80
∗17.29
0.79
0.57
0.98
0.29
5.51
4.64
6.82
5.29
5.38
6.39
5.15
6.48
5.80
6.27
4.51
5.55
4.37
4.09
4.72
4.56
0.35
0.22
0.27
0.20
0.12
0.05
0.05
0.01
0.07
0.08
0.07
0.09
0.12
0.17
0.09
0.14
60.29
60.12
60.54
60.09
100
100
100
100
17.30
0.66
5.56
5.85
5.53
4.44
0.26
0.06
0.08
0.13
60.26
100
15.85
15.51
15.65
15.92
16.21
16.01
2.16
2.49
2.44
1.63
1.60
1.72
6.19
5.17
4.92
6.00
6.03
5.87
5.07
5.60
5.91
5.09
4.91
5.15
4.58
6.03
5.82
6.01
5.87
5.90
5.14
4.41
4.47
4.60
4.46
4.52
0.30
0.30
0.31
0.30
0.32
0.31
0.07
0.04
0.04
0.05
0.07
0.05
0.07
0.07
0.08
0.06
0.06
0.07
0.09
0.13
0.13
0.12
0.12
0.12
60.48
60.24
60.22
60.22
60.34
60.28
100
100
100
100
100
100
15.86
2.01
5.70
5.29
5.70
4.60
0.31
0.05
0.07
0.12
60.30
100
Average
Kreep soils
12001,599
12033,464
14163,778
Bulk
Bulk
Bulk
Average
Low-Ti Mare soils
12001,599
15601
21000,5
24999,6
Bulk
Bulk
Bulk
Bulk
Average
High-Ti Mare soils
10084,1591
A17 drill core, Unit E
A17 drill core, Unit D
A17 drill core, Unit C
A17 drill core, Unit B
A17 drill core, Unit A
Bulk
1000–20 µm
1000–20 µm
1000–20 µm
1000–20 µm
1000–20 µm
Average
Note. Data were taken from Papike et al. (1982) and are reported in mol%. Original literature on Luna soil chemistry lacks in reporting values for Si (Laul and
Papike, 1981). Therefore, we assigned average wt% values of Si to empty Luna Si entries (marked by asterisk). All data are normalized to 100%.
3. The sputter model
where Eb is the surface binding energy of the sputtered particle
and Ec is the cut-off energy. The cut-off energy Ec , which is the
maximum energy that can be imparted to a sputtered particle
by a projectile particle with energy Ei , is given by the limit
imposed by a binary collision between a projectile atom M1
and the target atom M2 (to be sputtered) as
In the present model we only consider particle sputtering by
the impact of energetic ions, which in the case of the Moon is
the solar wind, on the planetary surface in a self-consistent way.
Particle sputtering will release all species from the surface into
space reproducing more or less the local surface composition on
an atomic level. Preferential sputtering of the different elements
of a compound will lead to a surface enrichment of those elements with low sputtering yields in the top-most atomic layers.
However, the steady-state composition of the flux of sputtered
atoms will reflect the average bulk composition. Thus, particle
sputtering, when operative, will give us compositional information about the refractory elements of the bulk surface.
The energy distribution for particles sputtered from a solid,
f (Ee ), with the energy Ee of the sputtered particle, has been
given as (Sigmund, 1969)
Ee
E e + Eb
6Eb
f (Ee ) =
(1)
1−
,
√
Ec
3 − 8 Eb /Ec (Ee + Eb )3
4M1 M2
(2)
.
(M1 + M2 )2
Fig. 1 shows examples of the energy distribution for typical
atoms sputtered from the lunar surface under solar wind bombardment. Note that the maximum of the energy distribution
(Eq. (1)) is at Emax = Eb /2, with Eb ranging from fractions
of an eV to several eV depending on species and matrix. At
higher energies the distribution falls off with Ee−2 , which is
also observed experimentally (e.g., Thompson et al., 1968;
Husinsky et al., 1985). Note that if the energy of the incident
ion is so low that the cut-off energy Ec approaches the binding energy the energy distribution of sputtered atoms becomes
rather narrow. Moreover, ion sputtering operates above an energy threshold of about 30 eV, which is exceeded significantly
by the solar wind ions.
The polar angle distribution of sputtered atoms, f (α), for
polycrystalline surfaces is best described by a quadratic angular dependence, f (α) ∝ cos2 α for laboratory experiments
(Hofer, 1991). However, Cassidy and Johnson (2005), found
sentative of the surface. This assumption may be inaccurate if
the finest fraction occupies a large fraction of the surface area
because the sputtering mechanism is only effective in the uppermost 100 nm of a solid.
Ec = E i
The lunar exosphere: The sputtering contribution
489
Fig. 1. Energy distribution for sputtered O, Si, Ca, and Fe atoms according to Eq. (1) using incident protons of 1 keV energy. The symbols indicate the escape energy
for these atoms, the vertical bars are only for better visibility.
recently that for the fine-grained and porous regolith a better
choice is f (α) = cos α, which is used in our calculations. For
the azimuth angle we used a uniform distribution over 2π . Having the energy, the azimuth and elevation angle we can calculate
all three components of the initial particle velocity v and the
trajectory of each sputtered particle. Using many such trajectories the vertical density profile Ni (h) can be calculated using
our Monte Carlo code (Wurz and Lammer, 2003). Either the
exospheric density at the surface or the column density can be
used for comparison with observational data.
The flux Φi of particles sputtered from the planetary surface
can be calculated as
Φi = Φion Yitot = Φion Yirel Ci ,
(3)
where Φion is the ion flux onto the surface and Yitot the total
sputter yield of species i from the surface, i.e., the elemental
mixture discussed above. The total sputter yield can be broken
up into a relative sputter yield Yirel and Ci the atomic abundance
of species i on the surface. The total sputter flux of species i is
also given by
Φi = Ni (0)vi (4)
with Ni (0) the exospheric particle density at the surface
(h = 0), and vi the average velocity of sputtered particles.
Combining Eqs. (3) and (4) we calculate the exospheric density
at the surface for species i as
Ni (0) = Φion Yitot
1
.
vi (5)
In the Monte Carlo calculation Ni (0) is used as a starting point
to calculate the density profile from the sputtering process for
a given surface composition. We can easily integrate the exospheric density profile to obtain the column density, which is
the typical measurement obtained from telescopic observations
of the exosphere.
The average release velocity is derived from the sputter distribution (see Eq. (1) and Fig. 1) as
vf (v) dv
vi = f (v) dv
3v12 + 5v22
1 2
3 arctan(v2 /v1 )
= v1 v2 − 2
(6)
+
2
v1 v2
(v1 + v22 )2
with the abbreviations
2Eb,i
4vSW
v1 =
and v2 =
,
m2
M1 + M2
where Eb,i is the binding energy of species i in the particular
chemical mix of the surface, m2 is the mass of the sputtered
atom in kg, and M1 and M2 are the atomic numbers for the incoming and the sputtered atom,
respectively. Note that the most
probable velocity is vmp = Eb,i /m2 , which is lower than the
average release speed by a factor of about 3.3. These velocities
have to be compared to the lunar escape speed of 2.376 km s−1 .
If we take oxygen as an example, with a binding energy of Eb =
2.0 eV, we get vi = 11.57 km s−1 and vmp = 3.47 km s−1 ,
which both exceed the escape velocity considerably. The same
is true for the other elements and is illustrated in Fig. 1. Thus
most of sputtered atoms escape the lunar gravity field.
In conclusion, if we know the flux of ions impinging on
the planetary surface, Φion , then with the sputter yields Yitot
we can calculate the sputtered flux, the surface density, the
density profile, and the column density ab initio and compare
these numbers with the observations using our Monte Carlo
code (Wurz and Lammer, 2003). The addition of sputter yields
together with a surface composition is the most significant improvement to the exosphere model reported earlier (Wurz and
Lammer, 2003). Such calculations for the Moon will be shown
in the next chapter.
The heavy ions in the solar wind are highly charged because
of the million-degree hot solar corona. Oxygen, for example,
is present in the solar wind with charges states of typically +6
490
P. Wurz et al. / Icarus 191 (2007) 486–496
and +7, iron with charge states in the range from +8 to +12.
These high charge states mean that these ions have high internal energies, e.g. 295 eV for O6+ and 1055 eV for Fe10+ , which
have to be compared to their kinetic energies in the solar wind
of typically 16 keV for oxygen and 56 keV for iron. It has been
argued that the sputter yield for highly charged ions impacting
on a planetary surface is increased by a factor of 10 to 1000
as a result of their high internal energy (Shemansky, 2003).
The laboratory data on sputter yields for highly charged ions
have been reviewed by Aumayr and Winter (2004), and we will
briefly summarize their findings here. For metallic surfaces and
semiconductors (Si and GaAs) no deviation of the sputter yield
for highly charged ions from the sputter yield of singly charged
ions was found, with the highest charge states investigated being Ar9+ and Xe25+ . Moreover, all the measured sputter yields
agree with the TRIM calculations, a software package which
considers only the kinetic energy of the impacting ion (Ziegler
and Biersack, 1985; Ziegler, 2004). For ionic crystals (NaCl
and LiF) a pronounced increase with ionic charge state was observed; for NaCl the sputter yield increased by a factor of 4
for Ar8+ ions compared to Ar+ ions, for LiF the sputter yield
increased by a factor of 25 for Ar14+ ions compared to Ar+
ions. Note that Ar charge states in the solar wind range from
+8 to +11. For oxides, which are the best analogue for the
lunar surface, a clear signature of a sputter yield increase for
highly charged ions was observed for SiO2 and Al2 O3 . For SiO2
this increase was about 3 for Ar8+ ions compared to Ar+ ions,
and about 65 for Xe25+ ions compared to Ar+ ions. Similar enhancements were found for the Al2 O3 surface. However, this
enhancement is strongly dependent upon the ion dose the surface has been exposed to. After a removal of about a monolayer
from the oxide surface, the sputter yield for highly charged ions
drops to about the values for singly charged ions. Removal of
a monolayer of surface material corresponds to a heavy ion
flux of a few 1013 ions cm−2 at solar wind energies, which
takes about two months at Earth orbit. This reduction in sputter yield is attributed to the top-most surface layers becoming
reasonably conductive (by preferential loss of oxygen and the
creation of crystal defects) and thus the highly charged ions become decharged, i.e., they lose their internal energy, when they
approach the surface. In conclusion, we do not consider an enhancement of the sputter yield by highly charged solar wind
ions in this paper.
4. Solar wind sputter yields for lunar soil material
We calculated the total sputter yield, Yitot , for all species i
for the four reference surface compositions discussed above
using the TRIM.SP software (Biersack and Eckstein, 1984;
Ziegler, 2004). The TRIM.SP software has been optimized to
and usually reproduces sputter yields very well (e.g., Gillen et al.,
2002; Aumayr and Winter, 2004). Also the energy distribution
of sputtered particles is reproduced well by TRIM.SP, even the
high energy tails (e.g., Gillen et al., 2002). In the sputter process
the energy of the impacting particle is distributed among atoms
at and near the surface in the form of a collision cascade, which
leads to the release of some of these atoms from the surface. The
flux of sputtered atoms is considered to be proportional to the
abundance of that element on the surface (see Eq. (3)), which
is used widely in laboratories for surface composition analysis.
Nevertheless, we calculated the individual sputter yields for all
elements for the four reference surfaces to avoid making this
assumption.
Depending on the abundance of a species in the lunar surface
between 75,000 and 1,700,000 projectile ions were simulated
with TRIM.SP to assure good statistics. As projectile ions we
used a mixture of protons and alpha particles, 95 and 5%, respectively. Heavy solar wind ions are not considered here since
their total abundances is about 0.1% in the solar wind (Wurz,
2005, and references therein). The solar wind velocities we investigated are in the range from 300 to 800 km s−1 to cover the
typical variation of solar wind conditions. The sputter yields
obtained from the TRIM.SP calculation for the four reference
surface compositions are shown in Fig. 2. We calculated different angles of incidence of the ions on the surface since the
sputter yield increases with shallower impact. The sputter yield
is a strong function of the incident angle of the ion with the
maximum of the sputter yield being between 55◦ and 85◦ (with
respect to the surface normal) depending and the mass of the
incident and the target atom (Sigmund, 1969, 1981). However,
this is only true for smooth surfaces. As soon as the surface
is rough (for example because of prolonged ion bombardment)
the angle of incidence dependence is gone and the sputter yield
has about the same value as for a smooth surface at 45◦ impact
(Küster et al., 2000). Since lunar regolith certainly can be regarded as a rough surface we used the sputter yields for 45◦ in
the following calculations.
For solar wind speeds other than the calculated ones, we numerically interpolate between the calculated values. Since the
velocity dependence of the sputter yields has a simple form, this
interpolation is straightforward and no significant uncertainty is
expected from this procedure.
The sputter yields presented in Fig. 2 are the yields for solid
grains. The porosity of the surface is not included at this stage,
but later in the calculation. As one can see from Fig. 2 the sputter yields are low, around 0.1 surface atom per solar wind ion.
The reason for the low sputter yields is that the solar wind is
mostly composed of protons; the 5% alpha particles typically
contribute 30% to the total sputter yield. Oxygen or iron ions of
the solar wind contribute at most 1% to the sputter yield, other
heavy ions even less. The velocity dependence is such that the
maximum of the sputter yield typically results from solar wind
speeds in the range between 300 and 500 km s−1 .
5. Exospheric densities from sputtering
Using the previously mentioned sputter yields and the energy and angular distributions for the sputtering process, we
calculated exospheric density profiles for the four categories
of lunar surface compositions. When calculating the sputter release for a space-weathered planetary surface one also has to
account for the porosity of the surface (Cassidy and Johnson,
2005). For lunar regolith breccias a typical porosity of 25 ± 7%
was found (Warren, 2001, and references therein). However,
The lunar exosphere: The sputtering contribution
491
Fig. 2. Sputter yields (i.e., sputtered atoms per incoming ion) for solar wind ions for the four compositional types of the lunar surface.
since sputtering probes the topmost surface it is more appropriate to use the porosity from cores, which is between 46 and 57%
(Costes et al., 1970). Note that porosity values up to 80% have
been reported (Kaula, 1968). In our calculations we use a porosity of 50%, which seems to be a good compromise for a global
characterization of the porosity of the uppermost lunar surface.
We studied three solar wind cases of solar wind speed
and flux: (1) vSW = 450 km s−1 , fp = 2.0 × 1012 m−2 s−1 ,
(2) vSW = 530 km s−1 , fp = 2.65 × 1012 m−2 s−1 , and (3) vSW
= 670 km s−1 , fp = 2.68 × 1012 m−2 s−1 , since these solar
wind parameters prevailed during the measurement campaigns
used to determine the upper limits given in Table 1 (based on
the OMNI data base). The calculations were performed for the
sub-solar point, for other points the sputter yields, and thus the
exosphere densities, will scale with the cosine of the zenith angle. The results of these calculations are shown in Fig. 3. Note
that the sputtering process releases surface atoms (and molecules) with high energies with respect to the lunar gravitational
field (see Eq. (1)). Practically all sputtered atoms escape from
the Moon.
In almost all cases the values predicted by our model are
consistent with the observational upper bounds shown in Table 1. The one exception is for the abundance of calcium atoms,
which is about the same and sometimes exceeding the 5-sigma
upper bound of 1 atom per cm3 of the measurement, in particular for the highland soils. However, it should be borne in mind
that the upper bound is derived from a spatially integrated measurement and is not appropriate for a pure highland soil, and
492
P. Wurz et al. / Icarus 191 (2007) 486–496
Fig. 3. Comparison of the calculations of the lunar exospheric surface densities at the sub-solar point with measurements (from Table 1). Note that the experimental
data are upper limits.
also that this is an unusually high solar wind speed. Moreover,
the calculated density is for the sub-solar point. If one assumes
a cosine law for the sputter yield because of the reduction of ion
flux with lower zenith angle, the integrated density value for the
sun-lit surface is a factor of 2π smaller and thus well inside with
the experimental upper bound. This apparent disagreement of
the model with the observational bounds is therefore insignificant. Clearly better observational data are needed in order to
test the assumptions and parameters of the model and this will
probably require in situ measurements.
Overall, the calculated exospheric densities for solar wind
sputtering are low, and the sum of all calculated elements gives
a surface density of about 1 × 107 m−3 . The total density of all
measured species is variable, but amounts to about 1012 m−3
on the dayside (Stern, 1999). Most of these species are volatiles
being released thermally or via PSD from the lunar surface and
thus have a low scale height in the atmosphere. This result is in
agreement with the observations of the lunar exosphere during
a lunar eclipse where the Moon was inside the Earth magnetosphere and sputtering by energetic ions should be a negligible
contribution to the lunar exosphere. But no variations in exospheric column contents at least for observable species like
Na were observed (Mendillo et al., 1999), and the authors concluded that solar wind sputtering may not be the main surface
release process which is in agreement with our results. However, in a later study a hysteresis effect in the sodium density
was observed, whereby the sodium density was lower coming
out of the magnetotail than it was coming in, which was interpreted that the solar wind causes a mobilization of the sodium
on the lunar surface (Potter et al., 2000).
The lunar exosphere: The sputtering contribution
493
Fig. 4. Left: density profiles of thermally released atoms and molecules based on the exospheric surface densities from Stern (1999). Right: density profiles of
sputtered atoms from Kreep soils for vSW = 670 km s−1 and fp = 2.85 × 1012 m−2 s−1 . Both calculations are done for the sub-solar point.
Sputtered atoms have more energy than thermally released
species and thus a larger scale height. Fig. 4 shows the density
profiles for sputtered atoms in comparison with thermally released species, in both cases for the sub-solar point. To model
the thermal release we assumed a surface temperature at the
sub-solar of 400 K and used our Monte Carlo code to calculate
the density profiles (Wurz and Lammer, 2003). For the thermally released species we used the surface densities for the
known species compiled by Stern (1999) as input for our calculation of the density profiles. Clearly, the volatiles dominate
the exosphere close to the lunar surface by several orders of
magnitude. Because of the much larger scale height of atoms
released via sputtering into the exosphere, they start to dominate the exosphere at altitudes exceeding at a few 1000 km,
with the exception of some light and abundant species released
thermally, H, H2 , and He. Because of their large densities CH4
and OH may also reach high altitudes. Note that the surface
density reported for OH is only an upper limit (Stern, 1999).
Moreover, these heavier hydrogen compounds could be ionized
by the solar radiation and particle environment and picked up
by the solar wind (Hodges, 1973).
One can see from our studies that the exospheric densities are very low, and thus it will be an experimental challenge to measure these densities accurately, especially from
an orbiting spacecraft. For the sputtered particles one can
also hope for transient events in the solar wind, i.e., coronal mass ejections (CMEs), during which the solar wind flux
can increase by one to two orders of magnitude in intensity and so will the exospheric densities. The duration of
a CME is typically a day at Earth orbit. In any case, sputtering, together with micro-meteorite impact vaporisation, are
the only means to bring refractory elements into the exosphere.
6. Solar wind induced composition changes on the surface
Since the Moon is unprotected from the solar wind by a magnetic field the total Sun-facing surface is constantly bombarded
by solar wind ions except when the Moon is inside the Earth’s
magnetotail. The total flux of solar wind ions onto the Moon’s
surface is typically 4.5 × 1012 ions m−2 s−1 , which is variable
with time. This ion flux corresponds to 8.5 × 10−15 kg m−2 s−1 .
Integrated over the lunar surface we get a total of 4.3 ×
1025 ions s−1 , which corresponds to 0.081 kg s−1 . When considering such a high influx of solar material we have to worry
if the composition of the topmost surface of the Moon may reflect the solar composition because of deposition (implantation)
of solar material. Fortunately, protons and alpha particles make
up more than 99% of the solar wind ions, and the lunar surface is saturated with hydrogen and helium. The heavy ions
(from carbon to iron and up) together are about 0.1% of the
solar wind ions in the number flux (Wurz, 2005, and references
therein). Although the sputter yield of solar wind ions is rather
low (about to 0.07 atoms/ion considering the mix of solar wind
ions and the porosity) it is more than sufficient to remove all
implanted heavy ions. For heavy atoms (from carbon to iron
and up), the removal via sputtering exceeds the input from solar wind by 1 to 3 orders of magnitude, because the abundance
of heavy ions in the solar wind is very low. Since the gravitational field of the Moon is low, practically all sputtered atoms
escape because their typical ejection speed exceeds the lunar
escape speed of 2376 m s−1 .
The meteoritic infall on the Moon is estimated to be 5.202 ×
10−16 kg m−2 s−1 (Bruno et al., 2006), which corresponds to
0.0197 kg s−1 or 5.68 × 1023 atoms s−1 for the full lunar surface. Meteoritic infall values of 6.3 × 10−16 kg m−2 s−1 and
a range of (4.8–6.3) × 10−16 kg m−2 s−1 the have been reported
494
P. Wurz et al. / Icarus 191 (2007) 486–496
earlier by Gault et al. (1972) and by Zook (1975), respectively.
The mass flux of meteoritic infall is low in comparison to the solar wind. Moreover, for meteorite infall the typical energy per
atom is much lower (typical impact velocities of 30 km s−1 )
than for solar wind. Thus, we conclude that on average the
release of particles into the exosphere via meteorite infall is
less important than by solar wind induced sputtering aside from
episodic events associated with the known meteoroid showers.
Interestingly, Zook (1975) estimated that the meteoritic fraction
in the released particles by an impact is about 20%.
In addition to loss of sputtering atoms from the exosphere,
ionization of exospheric species plays an additional role as an
escape process from the Moon. On the other hand it is expected
that 40% of the ionized exospheric particles may collide with
the surface (Manka and Michel, 1970) where they may also
act as sputter agents of the surface material or where they can
be implanted in the surface. Such a process is considered as
an explanation of the 40 Ar that is found in lunar soil samples (Manka and Michel, 1970; Hodges, 1975). Whether an ion
will be implanted in deeper surface layers of the lunar soil, or
will be adsorbed at the top surface layer or is released again,
depending on the impact energy (Bühler et al., 1966; Manka
and Michel, 1970). From sputter experiments with Ar ions it
is known that about 60% of the higher energy fraction of the
back-scattered ions with energies 1 < EAr 3 keV may be implanted in the surface at depths where they can not be ejected
by sputtering again.
For Ar ions with energies 3 keV almost 100% will be
trapped in the lunar grains (Bühler et al., 1966; Manka and
Michel, 1970). Manka and Michel (1970) found that depending on the IMF field strength, between 5–8% of the released
exospheric 40 Ar will be implanted in the surface material due
to ionization and subsequent interaction and acceleration by the
v × B force in the solar wind. A detailed study that will investigate the role of heavy exospheric ions to surface sputtering is
beyond the scope of this work but is considered for the future.
Generally, we expect that there is no large alteration of the lunar
surface composition because of the implantation of solar material. However, this will be different for heavier planetary bodies
without atmosphere, for example Mercury.
Hapke et al. (1975) conducted sputtering experiments on
artificial glass, whose chemical composition was similar to
Apollo 11 rocks, with the proton beam orientation at 45◦ from
the surface normal to simulate space weathering by ion bombardment. The forward sputtered atoms were captured by a Mo
foil substrate. The enrichment ratio (film/parent), measured by
means of electron microprobe analysis, gave clear evidence for
recondensation with increasing atomic weight of atoms on the
foil substrate, especially dominated by deposition of the element Fe. Hapke et al. (1975) and Hapke (2001, and references
therein) ascribe this phenomenon to be a key process in the
course of maturation and spectral reddening of surface soils
upon space weathering. Alkali elements were detected in the
film at very low levels indicating strong depletion in relation to
the parent, but no quantitative data were reported in Hapke et al.
(1975). However, since practically all sputtered atoms escape
the lunar gravitation field we conclude that selective reconden-
sation of atoms is not a relevant process for changing the surface
composition.
7. Conclusions
In almost all cases the values predicted by our model are
consistent with the observational upper bounds shown in Table 1. The only exception is the abundance of calcium atoms
for which the calculated density is commensurate with the 5sigma upper bound of 1 atom per cm3 as we discussed above.
The sum of all calculated elements released by the sputtering process gives a surface density of about 1 × 107 m−3
depending on solar wind parameters and surface composition. The total density of all measured species is variable, but
amounts to about 1012 m−3 on the dayside (Stern et al., 1997;
Stern, 1999). We conclude that at the surface the exosphere
is dominated by elements released via thermal processes and
PSD. Since both the solar wind and the solar illumination follow the same zenith angle dependence (neglecting the small
aberration of the solar wind because of the motion of the Earth
around the Sun) this conclusion is true for the entire dayside
of the Moon. Micro-meteorite impact vaporisation is considered to contribute to the exosphere even less than sputtering
but will be important on the night side of the Moon because
of the absence of the other processes. At higher altitudes the
sputtered contribution becomes more significant and eventually dominates, because of the low scale height of thermally
released particles.
Given the long range of released species on ballistic trajectories, both thermally released as well as sputtered species, smallscale fluctuations in the surface composition will be smeared
out with increasing altitudes. As a first-order approximation,
the radius of the footprint of an in situ measurement performed
at a certain altitude will be the same as the altitude. For the
Chandrayaan-1 mission orbiting at 100 km above the surface
only coarse compositional maps of the lunar surface can be
expected. However, we do expect to observe changes in the exospheric composition reflecting the varying surface.
We also conclude that the influx of solar wind and sputtering will not cause a change in the elemental composition of the
surface (other than the saturation of the surface with H and He)
because more heavy atoms are sputtered from the surface than
are delivered by the solar wind. Since most of the sputtered
atoms escape, selective recondensation cannot operate. However, the thermally released particles do not escape, with the
exception of H (63% escape), H2 (47% escape), and He (28%
escape), and are far more abundant near the surface. Their redistribution (day–night) and processing by UV light may cause
a change in the chemical composition of the surface.
Acknowledgments
The OMNI data were obtained from the GSFC/SPDF OMNIWeb interface at http://omniweb.gsfc.nasa.gov. The financial
support of the Swiss National Science Foundation is gratefully acknowledged. P. Dobnikar, C. Kolb, H. Lammer and
J.A. Martín-Fernández acknowledge support from the “Büro
The lunar exosphere: The sputtering contribution
495
für Akademische Kooperation und Mobilität” of the Austrian Academic Exchange Service under the ÖAD-Acciones
Integradas project 2006/2007. The chemical variability at planetary surfaces—implications for exogenic processes.
Hunten, D.M., Sprague, A.L., 1997. Origin and character of the lunar and mercurian atmospheres. Adv. Space Res. 19 (10), 1551–1560.
References
Johnson, F.S., 1971. Lunar atmosphere. Rev. Geophys. Space Phys. 9, 813–823.
Aumayr, F., Winter, H., 2004. Potential sputtering. Philos. Trans. R. Soc. London A 362, 77–102.
Bhardwaj, A., Barabash, S., Futaana, Y., Kazama, Y., Asamura, K., Sridharan,
R., Holmström, M., Wurz, P., Lundin, R., 2005. Low energy neutral atom
imaging on the Moon with the SARA Instrument aboard Chandrayaan-1
mission. J. Earth Syst. Sci. 114 (6), 749–760.
Biersack, J.P., Eckstein, W., 1984. Sputtering of solids with the Monte Carlo
program TRIM.SP. Appl. Phys. A 34, 73–94.
Bruno, M., Cremonese, G., Marchi, S., 2006. Neutral sodium atoms release
from the surface of the Moon induced by meteoroid impacts. Mon. Not. R.
Astron. Soc. 367, 1067–1071.
Bühler, F., Geiss, J., Meister, J., Eberhardt, P., Huneke, J.C., Signer, P., 1966.
Trapping of the solar wind in solids—I. Earth Planet. Sci. Lett. 1, 249–
255.
Cassidy, W., Johnson, R.E., 2005. Monte Carlo model of sputtering and other
ejection processes within a regolith. Icarus 176, 499–507.
Costes, N.C., Carrier, W.D., Mitchell, J.K., Scott, R.F., 1970. Apollo 11 soil
mechanics investigation. Science 167, 739–741.
Feastie, W.G., Feldman, P.D., Henry, R.C., Moos, H.W., Barth, C.A., Thomas,
G.E., Donahue, T.M., 1973. A search for far-ultraviolet emission from the
lunar atmosphere. Science 182 (11), 710–711.
Feldman, P.D., Morrison, D., 1991. The Apollo 17 ultraviolet spectrometer:
Lunar atmosphere measurements revisited. Geophys. Res. Lett. 18 (11),
2105–2109.
Flynn, B., Mendillo, M., 1993. A picture of the Moon’s atmosphere. Science 261, 184–186.
Flynn, B.C., Stern, S.A., 1996. A spectroscopic survey of metallic species abundances in the lunar atmosphere. Icarus 124, 530–536.
Gault, D.E., Hörz, F., Hartung, J.B., 1972. Effects of microcratering on the lunar
surface. Proc. Lunar Sci. Conf. 3, 2713–2734.
Gillen, D.R., Graham, W.G., Goelich, A., 2002. Sputtering of copper atoms
by keV atomic and molecular ions: A comparison of experiment with analytical and computer based models. Nucl. Instrum. Methods B 194, 409–
416.
Gorenstein, P., Golub, L., Bjorkholm, P., 1974a. Detection of radon at the edges
of lunar maria with the Apollo alpha-particle spectrometer. Science 183,
411–413.
Gorenstein, P., Golub, L., Bjorkholm, P., 1974b. Radon emanations from the
Moon, spatial and temporal variability. Earth Moon Planets 9, 129.
Hapke, B., 2001. Space weathering from Mercury to the asteroid belt. J. Geophys. Res. 106 (E5), 10039–10074.
Hapke, B., Cassidy, W., Wells, E., 1975. Effects of vapor-phase deposition
processes on the optical, chemical, and magnetic properties of the lunar
regolith. Moon 13, 339–353.
Heiken, G.H., Vaniman, D.T., French, B.M., 1991. Lunar Sourcebook: A User’s
Guide to the Moon. Cambridge Univ. Press, Cambridge, England, p. 736.
Hinton, F.L., Taeusch, D.R., 1964. Variation of the lunar atmosphere with the
strength of the solar wind. J. Geophys. Res. 69, 1341–1347.
Hodges Jr., R.R., 1973. Helium and hydrogen in the lunar atmosphere. J. Geophys. Res. 78, 8055–8064.
Hodges Jr., R.R., 1975. Formation of the lunar atmosphere. Moon 14, 139–157.
Hodges Jr., R.R., 1980. Methods for Monte Carlo simulation of the exospheres
of the Moon and Mercury. J. Geophys. Res. 85, 164–170.
Hodges Jr., R.R., Hoffman, J.H., 1975. Implications of atmospheric 40 Ar escape on the interior structure of the Moon. Proc. Lunar Sci. Conf. 6, 3039–
3047.
Hofer, W.O., 1991. Angular, energy, and mass distribution of sputtered particles. In: Behrisch, R., Wittmaack, K. (Eds.), Sputtering by Particle Bombardment. Springer, Berlin, pp. 15–90.
Husinsky, W., Girgis, I., Betz, G., 1985. Doppler shift laser fluorescence spectroscopy of sputtered and evaporated atoms under Ar+ bombardment.
J. Vac. Sci. Technol. B 3 (5), 1543–1545.
Kaula, W.M., 1968. An Introduction into Planetary Physics: The Terrestrial
Planets. Wiley, New York, p. 331.
Killen, R.M., Ip, W.-H., 1999. The surface-bounded atmospheres of Mercury
and the Moon. Rev. Geophys. 37 (3), 361–406.
Küster, M., Eckstein, W., Dose, V., Roth, J., 2000. The influence of surface
roughness on the angular dependence of the sputter yield. Nucl. Instrum.
Methods B 145, 320–331.
Laul, J.C., Papike, J.J., 1980. The lunar regolith: Comparative chemistry of the
Apollo sites. Proc. Lunar Sci. Conf. 11, 1307–1340.
Laul, J.C., Papike, J.J., 1981. Comparative chemistry of size fractions from the
Apollo and Luna sites. Meteoritics 16, 347.
Laul, J.C., Papike, J.J., Simon, S.B., 1982. The Apollo 14 regolith—Chemistry
of cores 14210/14211 and 14220 and soils 14141, 14148, and 14149. Proc.
Lunar Sci. Conf. 13, A247–A259.
Lawson, S.L., Feldman, W.C., Lawrence, D.J., Moore, K.R., Elphic, R.C.,
Belian, R.D., 2005. Recent outgassing from the lunar surface: The
Lunar Prospector alpha particle spectrometer. J. Geophys. Res. 110,
doi:10.1029/2005JE002433.
Manka, R.H., Michel, F.C., 1970. Lunar atmosphere as a source of Argon-40
and other lunar surface elements. Science 169, 278–280.
Mendillo, M., Baumgardner, J., Wilson, J., 1999. Observational test for the
solar wind sputtering origin of the Moon’s extended sodium atmosphere.
Icarus 137, 13–23.
Milillo, A., Wurz, P., Orsini, S., Delcourt, D., Kallio, E., Killen, R.M., Lammer, H., Massetti, S., Mura, A., Barabash, S., Cremonese, G., Daglis,
I.A., DeAngelis, E., Di Lellis, A.M., Livi, S., Mangano, V., Torkar, K.,
2005. Surface–exosphere–magnetosphere system of Mercury. Space Sci.
Rev. 117, 397–443.
Papike, J.J., Simon, S.B., Laul, J.C., 1982. The lunar regolith—Chemistry, mineralogy, and petrology. Rev. Geophys. Space Phys. 20, 761–826.
Papike, J.J., Ryder, G., Shearer, C.K., 1998. Lunar samples. Rev. Mineral. 36
(5), 1–234.
Potter, A.E., Morgan, T.H., 1988. Discovery of sodium and potassium vapor in
the atmosphere of the Moon. Science 241, 675–680.
Potter, A.E., Morgan, T.H., 1998. Coronagraphic observations of the lunar
sodium exosphere near the lunar surface. J. Geophys. Res. 103 (E4), 8581–
8586.
Potter, A.E., Killen, R.M., Morgan, T.H., 2000. Variation of lunar sodium
during passage of the Moon through the Earth’s magnetotail. J. Geophys.
Res. 105, 15073–15084.
Shemansky, D.E., 2003. The role of solar wind heavy ions in the space environment. AIP Conf. Proc. 663, 687–696.
Sigmund, P., 1969. Theory of sputtering. I. Sputtering yield of amorphous and
polycrystalline targets. Phys. Rev. 184, 383–416.
Sigmund, P., 1981. Sputtering by ion bombardment: Theoretical concepts. In:
Behrisch, R. (Ed.), Sputtering by Ion Bombardment, vol. I. Springer, Berlin,
pp. 9–71.
Sprague, A.L., Kozlowski, R.W., Hunten, D.M., Grosse, F.A., 1992. The
sodium and potassium atmosphere of the Moon and its interaction with the
surface. Icarus 96, 27–42.
Stern, S.A., 1999. The lunar atmosphere: History, status, current problems, and
context. Rev. Geophys. 37 (4), 453–491.
Stern, S.A., Parker, J.W., Morgan, T.H., Flynn, B.C., Hunten, D.M., Sprague,
A.L., Mendillo, M., Festou, M.C., 1997. An HST search for magnesium in
the lunar atmosphere. Icarus 127, 523–526.
496
P. Wurz et al. / Icarus 191 (2007) 486–496
Thompson, M.W., Farmery, B.W., Newson, P.A., 1968. A mechanical spectrometer for analyzing the energy distribution of sputtered atoms of copper and gold. Philos. Mag. 18 (152), 361–
383.
Warren, P.H., 2001. Porosities of lunar meteorites: Strength, porosity and petrologic screening during the meteorite delivery process. J. Geophys. Res. 106
(E5), 10101–10111.
Wieler, R., Kehm, K., Meshik, A.P., Hohenberg, C.M., 1996. Secular changes
in the xenon and krypton abundances in the solar wind recorded in single
lunar grains. Nature 382, 46–49.
Wurz, P., 2005. Solar wind composition. In: The Dynamic Sun: Challenges for
Theory and Observations, ESA SP-600, vol. 5.2. Noordwijk, The Netherlands, pp. 1–9.
Wurz, P., Lammer, H., 2003. Monte-Carlo simulation of Mercury’s exosphere.
Icarus 164 (1), 1–13.
Ziegler, J.F., 2004. SRIM-2003. Nucl. Instrum. Methods B 219, 1027–1036.
Ziegler, J.F., Biersack, J.P., 1985. The Stopping Power and Ranges of Ions in
Matter. Pergamon, New York. http://www.srim.org/.
Zook, H.A., 1975. The state of meteoritic material on the Moon. Proc. Lunar
Sci. Conf. 6, 1653–1672.